Soft-copy mammographic readings with different computer-assisted detectioncuing environments: Preliminary findings

Citation
B. Zheng et al., Soft-copy mammographic readings with different computer-assisted detectioncuing environments: Preliminary findings, RADIOLOGY, 221(3), 2001, pp. 633-640
Citations number
34
Language
INGLESE
art.tipo
Article
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging","Medical Research Diagnosis & Treatment
Journal title
RADIOLOGY
ISSN journal
0033-8419 → ACNP
Volume
221
Issue
3
Year of publication
2001
Pages
633 - 640
Database
ISI
SICI code
0033-8419(200112)221:3<633:SMRWDC>2.0.ZU;2-W
Abstract
PURPOSE: To assess the performance of radiologists in the detection of mass es and microcalcification clusters on digitized mammograms by using differe nt computer-assisted detection (CAD) cuing environments. MATERIALS AND METHODS: Two hundred nine digitized mammograms depicting 57 v erified masses and 38 microcalcification clusters in 85 positive and 35 neg ative cases were interpreted independently by seven radiologists using five display modes. Except for the first mode, for which no CAD results were pr ovided, suspicious regions identified with a CAD scheme were cued in all th e other modes by using a combination of two cuing sensitivities (90% and 50 %) and two false-positive rates (0.5 and 2.0 per image). A receiver operati ng characteristic study was performed by using soft-copy images. RESULTS: CAD cuing at 90% sensitivity and a rate of 0.5 false-positive regi on per image improved observer performance levels significantly (P < .01). As accuracy of CAD cuing decreased so did observer performances (P < .01). Cuing specificity affected mass detection more significantly, while cuing s ensitivity affected detection of microcalcification clusters more significa ntly (P < .01). Reduction of cuing sensitivity and specificity significantl y increased false-negative rates in noncued areas (P < .05). Trends were co nsistent for all observers. CONCLUSION: CAD systems have the potential to significantly improve diagnos tic performance in mammography. However, poorly performing schemes could ad versely affect observer performance in both cued and noncued areas.